Equilibrium adsorption on a random site surface

Equilibrium adsorption on a random site surface

J. Talbot 1 , G. Tarjus 2 and P. Viot 2
1 Department of Chemistry and Biochemistry, Duquesne University, Pittsburgh, PA 15282-1530
2 Laboratoire de Physique Theorique de la Matiere Condensee,
Universite Pierre et Mane Curie, 4, place Jussieu, 15252 Paris Cedex, 05 France

We examine the reversible adsorption of spherical solutes on a random site surface in which the
adsorption sites are uniformly and randomly distributed on a substrate. Each site can be occupied
by one solute provided that the nearest occupied site is at least one diameter away. The model is
characterized by the site density and the bulk phase activity of the adsorbate. We develop a general
statistical mechanical description of the model and we obtain exact expressions for the adsorption
isotherms in limiting cases of large and small activity and site density, particularly for the one
dimensional version of the model. We also propose approximate isotherms that interpolate between
the exact results. These theories are in good agreement with numerical simulations of the model in
two dimensions.

PACS numbers:

I. INTRODUCTION

Heterogeneity often plays an important role in various
adsorption processes and its presence may profoundly
modify the adsorption isotherms and other thermody-
namic properties compared to the homogeneous situa-
tion. Consequently, numerous articles [J 0], reviews [1,0]
and monographs 0, @ have been published that describe
experimental, theoretical and numerical studies in the
area. The disorder may originate from the energetic or
structural heterogeneity of the substrate or from the ad-
sorbate species. Heterogeneity may also result if the ad-
sorbed molecules are large enough so that multisite ad-
sorption becomes possible 0j-

In a recent article [|[ , we investigated the properties of
a lattice model of adsorption on a disordered substrate
that can be solved exactly (See also Refs.[9|, G3). We
showed that there exists an exact mapping to the system
without disorder in the limits of small and infinite activ-
ities and we exploited this result to obtain an approxi-
mate, but accurate description of the disordered system.

For continuous systems, structural disorder may be
represented by the random site model (RSM) in which
adsorption sites are uniformly and randomly distributed
on a plane [ill 0- The molecules, represented by hard
spheres, can bind to these immobile sites. Steric exclu-
sion is expressed by the fact that a site is available for
adsorption only if the nearest occupied site is at least one
particle diameter away. In addition, adsorption energy is
assumed equal for each adsorbed molecule. Therefore,
the disorder of this model is characterized by the dimen-
sionless site density, p Sl only. The degree of complex-
ity increases drastically because steric effects, which usu-
ally dominate the adsorption on continuous surfaces, are
modified by the local disordered structure of adsorption
sites. This model may be appropriate for the reversible
adsorption of proteins on disordered substrates (ljj, [Tij .

Oleyar and Talbot [12] examined the reversible adsorp-
tion of hard spheres on the 2D RSM and proposed an
approximate theory for the adsorption isotherms based

on a cluster expansion of the grand canonical partition
function. Although successful at low site densities, the
quality of the theory deteriorates rapidly with increasing
site density and fails completely above a certain density.

In sections II and III we develop a general statistical
mechanical description of the RSM model and we confirm
the intuitive result that in the limit of large site density
the system maps to hard spheres adsorbing on a continu-
ous surface. Then, for finite site density, we prove that for
the one-dimensional system in the limit of infinite activ-
ity, there is a mapping to a hard rod system at a pressure
[3P = p s . We also present an argument supporting the
validity of this result in higher dimensions. Note that
when the activity is "strictly" infinite, desorption is no
longer possible and the adsorption process is irreversible.
In this case the model has an exact mapping to the Ran-
dom Sequential Ad sorp tion (RSA) of hard particles on a
continuous surface [ilj .

In sections IV and V we propose approximate theoret-
ical schemes for the adsorption isotherms in one and two
dimensions that interpolate between the limits of small
and large activities for a given site density p s . Compari-
son with simulation results shows that these approaches
are a considerable improvement over the cluster expan-
sion.

For completeness we note that in addition to its appli-
cation to adsorption, the model is also interesting because
of its relationship to the vertex cover problem (la. [iH] . A
vertex cover of an undirected graph is a subset of the ver-
tices of the graph which contains at least one of the two
endpoints of each edge. In the vertex cover problem one
seeks the minimal vertex cover or the vertex cover of min-
imum size of the graph. This is an NP-complete problem
meaning that it is unlikely that there is an efficient algo-
rithm to solve it. The connection to the adsorption model
is made by associating a vertex with each adsorption site.
An edge is present between any two vertices (or sites)
if they are closer than the adsorbing particle diameter.
The minimal vertex cover corresponds to densest particle
packings. Weight and Hartmann[l5|, [l6| obtained an an-

2

alytical solution for the densest packing of hard spheres
on random graphs, but the existence of the geometry in
adsorption processes implies that the machinery devel-
oped to describe adsorption on random graphs cannot
be used in RSM models.

II. STATISTICAL MECHANICS OF THE
RANDOM SITE MODEL

The adsorption surface is generated by placing n s
points, representing adsorption sites, randomly and uni-
formly on a substrate, either a line in ID or a plane in
2D (the boundary conditions are irrelevant in the large n s
limit). Spheres of diameter a may bind, centered, on an
available adsorption site. A site is available if the nearest
occupied site is at least a distance a away. Two points
are connected, and therefore cannot be simultaneously
occupied, if they are closer than a.

The positions of the n s sites are denoted by Ri where
i is a index running from 1 to n s . The sites are quenched
during the adsorption-desorption process. The adsorbed
phase in equilibrium with a bulk phase containing ad-
sorbate at activity A can be formally described with the
grand canonical partition function:

s(a,{rj)=i+^- / ... /dr™n( i +/«')n^)

n=\ ' ^ ^ i>j i=l

where the microscopic density of sites r\ (r) is given by

» 7 (r)=£«(r-R i ),

(2)

A = exp(/3/z) is the activity, denotes the position of
sphere i and /y is the Mayer f-function which is equal
to —1 if spheres i and j are closer than a and other-
wise. Eq. (fl]) applies to a particular realization of the
quenched variables {Ri}. In order to average over disor-
der, it is necessary to take the average not of the partition
function, but of the logarithm of the partition function
By using standard rules of diagram theory [l8j|, one
obtains that

RSJ-E— / dv n U n {v 1 ,r 2 ,...,v n )J[^i) (3)
n=l n " J i=l

where the bar means that the average is taken over dis-
order, U n (ri, r 2 , r„) denotes the Ursell function as-
sociated with the Mayer f-functions of hard particles,

II, ,ii • fvM-

Let us denote the probability of finding adsorption sites
at positions Ri, i = 1, n s , as P(Ri, R2, ...R s ). We will
assume that the positions are uncorrelated so that

P(Ri,Ra,...R.) = fJP(Ri)

(4)

and will consider a Poissonian distribution of points,
P(R) = 1/A. The average of the site density ry(r) is
given by

V(r) =/-/ n dR * P ( R i' R2 '- R sM r )

J •* i=i

= Ps (5)

where p s = n s /A is the site density of the particles in a
system of area A (length in one dimension) .

We show in Appendix A that the average of the loga-
rithm of the partition function over disorder can be writ-
ten as

ln(S)=ln(H*(« = Ap a )) + 0(l/p,)

(6)

where

ln(S*(z)) = E^/ dr"t/„(n,r 2 ,...,r n ) (7)

i.e., ln(S*(z)), is the partition function of hard spheres on
a continuous surface at an activity z = Xp s . This result
shows that when the site density p s goes to infinity and
that the activity A goes to 0, with the constraint that
the product Xp s remains finite, the Random Site Model
maps to a system of hard particles in continuous space.

The number density of adsorbed molecules can be com-
puted directly from the partition function:

p{X > Ps) = A-(^

(8)

with again z = Xp s .

By using Eq. (|A8[) given in the Appendix, one obtains
to the second-order in l/p s

p(\,p s ) = p*{z)

P * {Z) dP * [z) '-0(1/ ^ (9)

Ps+p*(z) dz

where p*(z) is the number density of the hard sphere
model in continuous space. Since p*{z) is an increas-
ing function of the activity z, it is an upper bound for

The expansion of lnS and p(X,p s ) in powers of the
activity when A — > can also be generated from the
above equations in straightforward way.

III. THE LIMIT OF LARGE ACTIVITY

A. A mapping with the (homogeneous) equilibrium
hard sphere model

The limit of large activity, A — > 00, is expected to lead
to the maximum density of adsorbed spheres for a given
density p s of adsorption sites. Indeed, this limit com-
bines the presence of a relaxation mechanism, which is

3

induced by the infinitesimally small but non-zero desorp-
tion process that allows a sampling of hard-sphere con-
figurations, with the guarantee that no sites left open for
adsorption will stay empty. The one-dimensional model
is then amenable to an exact solution in the limit A — >• oo.
Interestingly, it maps onto an equilibrium system of hard
rods in continuum (ID) space at the same density (which
is of course a function of p s )

The adsorbed density p(p s , A — > oo) = p ma , x (p s ) in the
ID case can be obtained exactly with the following sim-
ple probabilistic argument. Assume that a given site is
occupied. Then all sites that lie within a distance a from
this site must be unoccupied. In the the optimally packed
system the next site beyond this must be occupied. The
average distance from an arbitrary point to the first site
is 1/ps giving the average distance between two occupied
sites as a + 1/ p s . The average coverage in the maximally
occupied system is thus

Pmax(ps) = — — • (10)

1 + Ps<J

In the following we set a = 1. When the site density is
low, p s << 1, one recovers the independent site approxi-
mation (Langmuir model), p max = Ps- As the site density
increases Eq. (TTU|) shows that there is a continuous in-
crease of /O max and that a closed packed configuration is
obtained as p s approaches infinity

The above argument can be generalized to describe
the correlation functions, or more conveniently in this
ID system, the gap distribution functions. Specifically
let F{x) denote the probability density associated with
finding a gap of size between x and x+dx. In an optimally
packed configuration, the probability to find a gap of
length x is related to the probability to find the first site
at a given distance x (say to the right) of an arbitrary
point. For a Poissonian distribution of sites this simply
leads to

F max (x-p s )=p s e- p ° x . (11)

The reasoning is easily extended to multi-gap distribu-
tion functions, F[x\, x%), F(x±, X2, £3), where two suc-
cessive gaps are neighbors in the sense that they are sep-
arated by a single particle. One then shows that all
these higher-order gap functions factorize in products
of one-gap distribution functions, e.g. F max (xi, x%) —
-Pmax(a^i)-Fmax(^2). The outcome of this probabilistic ar-
gument is that the ID random site model is equivalent
to an equilibrium system of hard rods on a (continuous)
line at the pressure

0P = Ps- (12)

Indeed, from the known equation of state of the hard rod
fluid Hfll one has (3P = p/(l - p), i.e. p = @P/(1 + (3P),
which corresponds to Eq. (flU)) after insertion of Eg. (fT2l -.
and F(x) = (3Pe~P Px , which reduces again to Eq. (fTTjl .
The multi-gap distribution functions are also given by
products of 1-gap distribution functions, which completes
the proof of equivalence.

Extension of the above arguments to higher dimensions
is far from straightforward. First, in d — 2 and higher,
one may encounter at high adsorbed density phase tran-
sitions to ordered, crystalline-like, phases. Second, the
simplicity of the reasoning in terms of gaps character-
ized only by their length is lost when one leaves the one-
dimensional case. For these reasons, we have not been
able to develop a rigorous demonstration of a mapping
between the 2D RSM and the (homogeneous) hard-disk
system at equilibrium at the same density when the ac-
tivity A goes to infinity.

To nonetheless make some progress, let us consider the
nearest neighbor radial distribution function H(r) intro-
duced by Torquato |2l]. H(r) (and also used in the con-
text of irreversible adsorption models [22|,[23[ is the prob-
ability density associated with finding a nearest neighbor
particle center at some radial distance r from the refer-
ence particle center. This somewhat generalizes the 1-
gap distribution function to any spatial dimension. For
a Poissonian distribution of particle centers of density p s
in d — 2, one finds that

H(r) = 2irrp s e- 7Tr2p ° (13)

No exact formula exists for H(r) in a hard disk fluid. Us-
ing the results of reference [21| one can, however, demon-
strate that in the large r limit at an equilibrium pressure
P it is given by

H{r) ~ 2Trrf3Pe-^ Tr2 > 3P (14)

The reasoning now goes as follows. Consider a maxi-
mally occupied configuration (associated with the limit
A — > 00) in the 2D RSM with an adsorption site density
p s and consider a given adsorbed particle. All sites within
a radial distance a of the central occupied site must be
unoccupied. H(r) is associated with situations such that
the nearest adsorbed particle center is at a radial dis-
tance r of the reference one. Therefore, there should be
no adsorption sites in the region delimited by the two
circles of radius a (inside) and r (outside). Actually, this
statement is not quite right: in 2 dimensions, there may
be an exclusion effect due to other particle centers at a
distance > r of the central site. To be more rigorous, the
outside circle delimiting the region with no adsorption
sites should be (at least) of radius r — a. When r — ► 00,
one can neglect the contribution of width a due either to
the disk of radius a centered on the reference site or to
the shell of width a located between r — a and r. When
r — > 00, the probability that, given that a reference par-
ticle is centered at the origin, a spherical region of radius
r is empty of adsorption sites is given by exp(— p s 7rr 2 ),
so that asymptotically H max (r) goes as in Eq. (p~3|) . i.e.

ffm«(r) ~ 2-Krpse-^P' (15)

Comparison with Eq. (TT4"|) tells us that it is the same
asymptotic behavior as that of an equilibrium hard disk
system in the plane with /3P = p s . This is the same

4

1/ln(X)

FIG. 1: Simulated adsorption Isotherms. p s = 1,2,4 top to
bottom. The dashed lines show the predictions of Eq. (|10|l .

result as found in d = 1 (see above), except that the
density p and pressure P are no longer related by a simple
analytical expression.

B. Numerical verification of the suggested mapping

We confirmed Eq. (JTUJ) in two ways. The first generates
configurations of maximum density directly. A number
n s points are distributed uniformly and randomly in the
unit interval. The first site is occupied by a rod of length
a = l/n s . Each site to the right is checked in order until
one is found that is at least a distance a from the occu-
pied one. This site is occupied and the process iterated
until all sites have been accounted for. A number of av-
erages over different configurations of sites is performed.
We also verified Eq. (fT0|) by determining the adsorption
isotherms for different values of the site density p s and
taking the limit A — > oo: See Fig. [TJ

For the hard rod fluid at equilibrium the gap distribu-
tion function is given exactly by:

F(p,x) = -^ex V (--^). (16)
1-p 1 - p

The gap distribution function calculated for the dens-
est configurations of the random site model is in agree-
ment with the predictions of Eq. (|16p using a density
computed from Eq. (JTUJ), confirming that the configu-
ration of rods on the random site surface has the same
structure as the equilibrium hard rod fluid at the same
density (see FigJ^]). This is consistent with the behavior
of a related lattice model for which we showed that there

FIG. 2: Gap distribution function F ma x(a:) for configurations
of hard rods at infinite activity on the random site surface
with p s = 0.7,1,2.0,4.0 from right to left, bottom. The
dashed lines are the predictions of Eq.fTTJ The solid lines
correspond to the simulation results.

is an exact mapping to the system without disorder at
infinite activity [8j.

In order to use Eq. (|T5|) to estimate the maximum cov-
erage of the RSM in two-dimensions, we combine it with
the approximate equation of state for hard disks on a
continuous surface, which was proposed by Wang [241]:

— = 1+ ; (D+2a9+4b9 2 +8c9 3 +16d9 i +32e9 5 )

p 1-0/00

(17)

where 9 = Tia 2 p/A is the coverage, 9$ — 0.907.. is
the coverage of a hexagonal close-packed configuration,
D = 4.08768, a = 1.25366, b = 0.46051, c = 0.152797, d =
0.04412, e = 0.00929.. For a given value of the site den-
sity, p s , Eq. (|17p is solved numerically for 9. It is conve-
nient to introduce the dimensionless site density

a = ^a 2 p s (18)

The results, shown in Fig. [3] are in excellent agreement
with the simulation results for the entire range of a. This
should be compared with cluster expansion to second or-
der that gives good predictions only for a < 0.3[12]. Note
that the approximate equation of state, Eq. (|17p . does not
include the possible presence of a phase transition.

5

be expressed as

FIG. 3: Maximum coverage of the RSM as a function of the
dimensionless site density. The symbols show the simulation
results, the dashed line is the cluster approximation to second
order and the solid line shows the predictions of Eqs. (|12|)
and (fT7)) .

InS = Ni In Hi + N 2 In S 2 + N 3a In E a + N 3b In E 3b +

(19)

where N — XiN s is the number of clusters of type i
and i = (n, a) with n being the number of sites in the
cluster and a characterizing when necessary the subclass
of clusters (see FigQJ. This simple expansion is possible
because adsorption on a given cluster does not affect any
of the others. The adsorption isotherm is then

pA = XN X

which gives

In Si

XN 2

d In Ej
~0X~

(20)

P = \p s (xi

3 + 2A
+ 1 + 3A + A 2

1 + A

X 3a

1 + 2A
3

1 + 3A

3'2
X 3 b

(21)

In one-dimension, exact expressions can be obtained
for the first few clusters following the approach of Quin-
tanilla and Torquato [25[:

1+X

1+3X+X 1

A

1+2X

1+3X

FIG. 4: Lowest order clusters of adsorption sites. A solid line
connects two sites that are closer than a and a dashed line
indicates that two sites are further apart than a. Triplets of
type "3a" (left) and "3b" (right) are shown left to right in
the second row. The expression to the right of the clusters is
the corresponding grand canonical partition function for the
adsorbed particles.

IV. RSM IN ONE DIMENSION
A. Low site density expansion

To provide a description of the adsorption isotherms
at finite activity, it is useful to consider an expansion
in the density of adsorption sites. Assuming that the
adsorption surface consists of isolated clusters of sites,
the (averaged) logarithm of the partition function may

x x = exp(-2p s ) (22)

x 2 = exp(-2p s )(l - exp(-p s )) (23)

x 3a = cxp(-3 / 9 s )(exp(- / 9 ;5 ) + p s - 1) (24)

x 3b = exp(-2j(? s )(l - (1 + p a ) exp(-p s )) (25)

where x 2 , for example, is the number of clusters with
two linked sites (per adsorption site). At the triplet level
we need to distinguish between two subclasses of clusters
shown in Fig. |H In type "3a" two of the sites can be
simultaneously occupied, while in type "3b" only one of
the sites can be occupied since all are mutually closer
than the particle diameter.

The predictions of Eq. (|2"Tj) are compared with the sim-
ulation results in Fig. [5l The expansion to third order
provides a good description of the isotherm for p s = 0.2,
but the quality deteriorates rapidly for larger site den-
sity. Eq. (|2ip is unable to predict the correct limit when
A — > oo. This disagreement is more pronounced when p s
increases: for instance, when p s = 1, the expansion fails
completely, because the predicted density is lower than
for p s = 0.5.

Since the above expansion is limited to small values of
p s , we simplify Eq. (|21[) by performing a series expansion
in p s , giving

2A

+ A (1 + A)(1 + 2A)'
A 2 (12A 2 + 29A + 9)

2(1 + A)(l + 2A)(1 + 3A)(1 + 3A + A 2 ) '

0(p 3 s ))

(26)

6

We note that Eq. (|26|) can be expressed as

p(Ps, A) =

1=1

1 + A

+ F l (p s ,X)\ p[ (27)

where Fi(p s , A) represents the remaining terms in the ex-
act expansion and, consistent with these terms, has the
property that Fi(p s ,X) — > when A — > and A — > cxx
This property has been verified order by order for a ran-
dom lattice model[§j, and only for the three first orders
for this model. If we partially resum the series, neglecting
the remaining terms, we obtain that

Xp s

1 + A(l + Ps )

(28)

Although the site density expansion, Eq. (I26|) . is valid
only for small p s , the partial resummation leads to an
expression for p that is exact in the limit of very large
activities A, whatever the value of p s . It thus provides a
sensible approximation based on the site density expan-

Isotherms given by Eq. (f2"8")) are plotted in Fig. [5] (dot-
ted curves). Although for very low site density the third
order expansion gives a better estimation, the quality of
Eq. P5|) is significantly better for p s — 0.5, and it always
gives the saturation density exactly. For larger values of
the site density, p s > 1, the approximation overestimates
the adsorbed amount for intermediate activities. This
is expected since the neglected terms are non-zero for
intermediate bulk activity. Attempts to obtain a bet-
ter resummation of the series expansion, Eq. (I26|) . were
fruitless. Therefore, we follow the approach that we used
in our study of the model of dimer adsorption Q, by in-
troducing an effective activity.

B. Effective activity approach

Despite the exact mapping found in the infinite ac-
tivity limit, no simple reasoning can be used to obtain
exact results at finite bulk activity. We therefore investi-
gated several approximate methods, the most successful
of which is based on an effective activity.

The equation of state of the hard rod fluid on a con-
tinuous line is [20(

PP =

1 — per

(29)

According to the results in Section III, the hard rod fluid
at a pressure f3P = p s has the same density (obtained by
inverting Eq. (|2"9"]) ) as the densest configuration of hard
spheres on the random site surface, Eq. (fTU)) . We seek a
generalization of this mapping for an arbitrary value of
the bulk phase activity, i.e. an effective activity A e fi such
that the density of adsorbed rods in the RSM, p(p s , A) is
given by

ln( X)

FIG. 5: Adsorption isotherms versus activity A for several site
densities: p 3 = 0.5, 0.3, 0.2 from top to bottom. The dash-
dotted and dotted lines show the predictions of Eq. [21] (exact
expansion to third order in p s ) and Eq. 1281 respectively. The
solid lines show the simulation results.

where p*(X) is the density of rods on a continuous sub-
strate at an activity A. This is given exactly by

P*(A)

MA)

i+L w {\y

(31)

where L w (x) , the Lambert-W function, is the solution of
x = L w (x) exp(L w (x)). The inverse relation is

A =

P

; exp( —

(32)

From the exact result, Eq. (fTTJ|) , we have that A e ff(A =
oo) = p s e Ps . Taking Eqs. (|31[) as a mere definition of
A e ff(A,p s ), we have computed it from the simulated ad-
sorption isotherms. Results are shown in Figs. [5][5] and
will be used as a reference for testing the validity of var-
ious approximations.

Following the method introduced in the study of the
lattice modelfl], we attempt to construct simple expres-
sions that can reproduce the numerical results, when
available and describe the isotherms for a wide range of
parameters p s and A. The effective activity as a function
of the activity A and of the density site p s can be written
as

1 = f(Ps,X) | 1
A e ff p s A p s eP°

(33)

p(p s ,X) = p*(X cS (X,p s )),

(30)

where f{p s ,X) is a function to be determined. We
know that at small A, the exact behavior is given by

7

200

150

\a(X)

100

10 12

FIG. 6: Effective activity A e a as a function of the logarithm
of the activity In A with p s = 1. The solid line shows the
simulation result, calculated from the density by using Eq.
(1321) . The predictions of Eq. J33} combined with Eq. (J34J) and
Eq. (l36|l are shown as dotted and dashed lines, respectively.
They are almost indistinguishable. The horizontal dashed
line shows the exact limiting value of p s e Ps ■

10 12

FIG. 7: Same as Fig. [6] except that p s = 2. The prediction
of Eq. (|36p is now more accurate than that of Eq. (|34|l .

A c ff = p s A, which imposes that f(p s , A) goes to 1. Con-
versely, when A is large, the effective activity A e ff be-
haves as p s e Ps + O(A), which leads to the constraint that
liniA^oo f(Psi A) = A(p s ) where A(p s ) is an unknown
function of p s . If A(p s ) remains close to one, the sim-
plest choice consists of choosing

/(p s ,A) = l (34)

for all values of A. While this simple choice gives a

FIG. 8: Same as Fig. except that p s = 4. Eq. (J34} fails to
reproduce the simulation data for intermediate value of ln(A)
whereas the prediction of Eq. (|36p remains accurate.

fair agreement with the simulated values when p s = 1
(see Fig. the situation deteriorates for larger values
of p s (see Figs. [?)■ [8]). However, contrary to the one-
dimensional lattice model, we do not know the exact
asymptotic behavior of f{p s ,\) for large A. In order
to improve the description, we consider a homographic
function of A verifying the exact behavior, when A — *
and A — > oo,

f(Ps,X)

A(p s )X + B( Ps )
A + B(p s ),

(35)

where A{p s ) and B{p s ) are unknown functions. Exami-
nation of the series expansion in the limit of infinite bulk
activity suggests that B(p s ) should grow as e Ps in order
to give the right behavior. In the absence of additional
criteria, we have chosen A(p s ) — p s and B(p s ) = p s e p " ,
which gives

/(P., A)

A)

A + p s

(36)

Note that when p s = 1, this function gives f(p s ,X) = 1.
Figs. [BJ- H] show that the quality of the approximation
given by Eq. (f3"3")) and Eq. (fM)) deteriorates with in-
creasing p s . This is a consequence of the fact that the
asymptotic approach to the saturated state is not cor-
rectly captured with this simple approximation. A better
agreement with simulations is obtained by using Eq. (|33|)
and Eq.® for p s = 2 and p s = 4 (see Figs. [7] and HJ.

The purpose of this exercise is ultimately to obtain an
approximate, but accurate, description of the adsorption
isotherms. This is done by substituting Eq. (|33p in Eq.
(l3"Tj) . Fig. shows a comparison of these estimates with
the simulation results. The effective activity approach is

8

10 12

0.01

FIG. 9: Adsorption isotherms versus activity A for several
site densities: p s = 1,2,4, predicted by the effective activity
approach, Eqs. (|32|l and (|31[1 by using Eq. (|34p (dotted lines)
and combined with Eq. (|36[) (dashed lines). The solid lines
show the simulation results.

a considerable improvement over the cluster expansion,
whose accuracy is restricted to small density site p s < 0.5
(c.f. Fig. [5]) and allows one to describe adsorption even
when p s > 1.

C. Structure

We have also examined the structure of the hard-rod
configurations at finite activity. As in the case of the lat-
tice modelQ, we do not expect an exact mapping to the
homogeneous system at the same density. Nevertheless,
as shown in Fig. 1 1 01 the distributions computed from
the simulation are very well described by Eq. (|16[) with
a density equal to the equilibrium isotherm value.

V. TWO-DIMENSIONAL MODEL

We construct approximate isotherms in the same way
as for one-dimensional model, i.e. by introducing an ef-
fective activity as a function of bulk activity.

We first consider the homogeneous hard-disk system.
We determine the Helmholtz free energy per particle by
integration along the isotherms:

(37)

The excess chemical potential can then be obtained from

FIG. 10: Gap distribution function for configurations of hard
rods at finite activity on the random site surface with p 3 — 10.
The dashed lines are the predictions of Eq. [16] with a density
given by the equilibrium isotherm value and the solid lines
show the simulation results. A = 1,5,50,1000 from left to
right in the bottom part.

By inserting the approximate equation of state, Eq. (JT7J)
in Eq. (|3"7) . and substituting the result in Eq. ([581) . one
obtains for the activity, A = exp(/3/i) the relation

OR

ln(A) = ln(40/7r) e(9 5 - 20 d<9 4 c6 3 - 6 b9 2 - 4 at

5 3

-Din 1 -

2 V3\

DV36*
7T - 2 V30 '
(39)

where, we recall, 9 = ira 2 p/4.

The isotherm for the 2D RSM can be calculated by
following a procedure similar to that used in the one-
dimensional case. By means of Eq. (fT2"|) , one obtains the
saturation coverage when the bulk activity is infinite, #oo .
Then, by inserting this quantity in Eq. (|39|) , the effective
activity A^ = A e ff(#oo) is obtained.

The counterpart of Eq. (|33[) for the two-dimensional
system is now

1

Acff

1

1

A off

(40)

j3pL ex = (3f ex + i

de

(38)

where, in the absence of more information, we have set
/(P.,A) = 1.

Therefore, at a given bulk activity A and site density
p s , one obtains an effective activity by using Eq. (|4"0")l .
and by inserting the effective activity in Eq. (IST))) . one
deduces the corresponding coverage in the 2D RSM.

9

FIG. 11: Adsorption Isotherms for dimensionless site densities
a = 1,2,4 from bottom to top. The dashed lines show the
predictions of Eg. (|39p combined with Eg. (|40p . The solid lines
show the simulation results obtained with 20 realizations of
n s = 2000 sites.

Fig. [TT] compares the simulation results and the ap-
proximate isotherms for different values of the site den-
sity. While the agreement is not perfect, the scheme
represents a major improvement compared to the pertu-
bative approach (third-order density expansion) for de-
scribing situations where the density site is moderate to
high and allows one to develop new approximate isotherm
equations.

VI. CONCLUSION

We have presented a theoretical and numerical study
of the reversible adsorption of hard spheres on randomly
distributed adsorption sites. Unlike the case of irre-
versible adsorption, there is no simple mapping between
the RSM and a system of hard spheres on a homoge-
neous surface. We have, nonetheless, been able to ob-
tain some exact results in various limits: small and large
bulk activity and small and large site density. We have
proposed an effective activity approach, which interpo-
lates between the known behavior, to obtain an approx-
imate description for intermediate situations. The ad-
sorption isotherms predicted by this approach are in ex-
cellent agreement with simulation results. In the two-
dimensional case, we have not investigated phase transi-
tions in detail, although there does appear to be a solid-
like phase for sufficiently high site density and activity.
It will also be interesting to investigate reversible ad-
sorption on a substrate when a second source of disorder

is present: distribution of adsorption site energies. The
present approach should apply to this case as well.

APPENDIX A: RANDOM SITE MODEL:
BEYOND THE CONTINUUM LIMIT

The sites being uncorrelated, Eq.Q, one eas-
ily obtains the connected n-site correlated functions

^(ri)7/(r 2 )...r?(r„)^ which read

(p{riHr2)-v(r n )) c = Ps Jpfa - ( A1 )

The n-site correlation functions which appear in Eq. |T])
can be obtained from the above connected functions by
using an expansion in the (mean) site density:

77(n)77(r 2 )...77(r n ) = p'l + pT 2 X (WM^f) c

i<j

i<j<k
i<j<k<l

i<3

i<k<l

(A2)

The expression of the higher-order terms rapidly becomes
tedious.

On the other hand, the n-body connected density func-
tions of a homogeneous hard sphere system at activity z
are given by functional derivatives of the the logarithm
of grand-canonical function (see Eq.(j3|)

pW*(r ir . 2 ...r p |*

SP ln(S*(z(r)))

z{r)—z

X

(n — p)\

n—p

8z(yi)8z(y 2 )...8z(yp)
J dr p+1 ...dr n U n (r 1 ,r 2 ...r p ,r p+1 ...r n ).

(A3)

By inserting Eq. (jA2| in Eq.© and by using Eq.fA3}, the
disorder averaged logarithm of the partition function of
the RSM has the following expansion:

ni^ = ln(H*(^) + ^-|dr 1 p( 2 )*(r 1 ,r 1 |^

+ 3!7 y*ipi 3) *(n,ri,nW

+ ^3 J *i^ 4) *(ri,ri,ri,ri|z)

+ 4^2 J dridr 2 ,9* (4) (ri,ri,r2,r2|z)

+ ... (A4)

10

For hard spheres, the non-overlapping property leads
to a simple expression for the n-body connected density
function

(rx.n.^n.nl^) = {-if^ip- i)\p*(z) p . (A5)

A straightforward but more tedious calculation allows
one to obtain

p( 4 )*(r 1 ,r 1 ,r 2 ,r 2 |z) = V(z) 2 (0 ( 2 )*(r 1 ,r 2 )-2p( 2 )*(r 1 ,r 2 ) 2

(A6)

By inserting Eqs. (IA5j) and (IA6j) in Eq. (IA4j) . one obtains
that

MS) _ ln(3*(z)) , ( p*(zY , p*{zY p*(z)

A
p*{zf

2p:

A ' \ 2 Ps 3 P 2 Apl
2 dr(2h*(r\z) - (h*(r\z))) 2 (A7)

where h*(r\z) is the radial pair correlation correlation of
the hard-sphere model at the activity z[l8[. It is worth
mentioning that the second term of the rhs of Eq . (|A4j)
corresponds to successive terms of a series expansion that
can be resummed, and finally, one obtains that

^ = ^M + , s lnfl + ^]-^)

dv{2h*{r\z) - (h*(r\z)) 2 ) + 0(l/p 3 s ). (A8)

V. A. Bakaev, Surface Science 564, 108 (2004).

Z. Adamczyk, B. Siwek, P. Weronski, and E. Musial, Ap- [14

plied Surface Science 196, 250 (2002).

A. Dabrowski, Adv Colloid Int Sci 93, 135 (2001). [15
Z. Adamczyk, K. Jaszczolt, A. Michna, B. Siwek,
L. Szyk-Warszynska, and M. Zembala, Adv Colloid Int [16
Sci 118, 25 (2005), ISSN 0001-8686.

M. Jaroniec and R. Madey, Physical adsorption on het- [17
erogeneous solids (Elsevier, 1988).

W.Rudzinski and D. H. Everett, Adsorption of Gases on [18
Heterogeneous Surfaces (Academic Press, 1992).
R. D. Johnson, Z. G. Wang, and F. H. Arnold, J. Phys. [19
Chem. 100, 5134 (1996).

J. Talbot, G. Tarjus, and P. Viot, Phys. Rev. E 76, [20
051106 (2007). [21
G. Oshanin, O. Benichou, and A. Blumen, J. Stat. Phys. [22
112, 541 (2003).

G. Oshanin, O. Benichou, and A. Blumen, Europhys. [23
Lett 62, 69 (2003).

X. Jin, N. H. L. Wang, G. Tarjus, and J. Talbot, J. Phys. [24
Chem. 97, 4256 (1993). [25
C. Oleyar and J. Talbot, Physica A 376, 27 (2007).
E. Ostuni, B. A. Gryzbowski, M. Mrksich, C. S. Roberts,

and G. M. Whitesides, Langmuir 19, 1861 (2003).

E. Liithgens and A. Janshoff, ChemPhysChem 6, 444

(2005).

M. Weight and A. K. Hartmann, Phys. Rev. Lett. 26,
6118 (2000).

M. Weight and A. K. Hartmann, Phys. Rev. E 63, 056127
(2001).

M. L. Rosinberg, G. Stell, and G. Tarjus, J. Chem. Phys.
100, 5172 (1994).

J.-P. Hansen and I. R. McDonald, Theory of Simple Liq-
uids (Academic Press, 1976).

G. Stell, Phase transitions and critical phenomena (Aca-
demic Press, 1976), p. 205.
L. Tonks, Phys. Rev. E 50, 955 (1936).
S. Torquato, Phys. Rev. E 51, 3170 (1995).
M. D. Rintoul, S. Torquato, and G. Tarjus, Phys. Rev.
E 53, 450 (1996).

P. Viot, P. R. V. Tassel, and J. Talbot, Phys. Rev. E 57,
1661 (1998).

X. Z. Wang, Phys. Rev. E 66, 031203 (2002).

J. Quintanilla and S. Torquato, Phys. Rev. E 54, 5331

(1996).